Anal. Chem. 2005, 77, 8050-8056
Single-Cell Microarray for Analyzing Cellular Response Shohei Yamamura,†,‡,§ Hiroyuki Kishi,§,| Yoshiharu Tokimitsu,| Sachiko Kondo,| Ritsu Honda,| Sathuluri Ramachandra Rao,†,‡ Masahiro Omori,‡ Eiichi Tamiya,*,‡ and Atsushi Muraguchi|
Toyama New Industry Organization, 529, Takada, Toyama 930-0866, Japan, School of Materials Science, Japan Advanced Institute of Science and Technology, 1-1, Asahidai, Nomi City, Ishikawa 923-1292, Japan, and Department of Immunology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, 2630, Sugitani, Toyama 930-0194, Japan
Detection of cellular response by measuring intracellular calcium, (Ca2+)i with Ca2+-dependent fluorescent dye are standard approaches to detect ligand-stimulated cells and to study signaling through ligand/receptor interaction. We describe a single-cell microarray system to analyze cellular response of individual cells such as lymphocytes using microchamber array chips. The single-cell microarray chip is made from polystyrene with over 30 000 microchambers, which can accommodate only single cells. Lymphocytes derived from mouse spleen or human blood were spread on the microarray, and over 80% of the microchambers achieved single-cell status. Stimulation of B-cells through antigen receptors on the microarray allowed us to detect activated B-cells by comparing the states of single B-cells before and after stimulation with antigen, which is disabled for flow cytometry. In addition, this novel method demonstrated retrieval of positive single B-cells from microchambers by a micromanipulator and achieved antibody DNA analysis. The system is suitable for highthroughput analysis of intracellular Ca2+ response at the single-cell level and is applicable to screen antigen-specific lymphocytes for making specific monoclonal antibody. Alteration of intracellular calcium, (Ca2+)i, is a key event in signal transduction in cells stimulated through ligand/receptor interactions. Ligands bind to receptors, activating various tyrosine or serine/threonine kinases; the following activation of phospholipase C converts phosphatidylinositol to inositoltrisphosphate, which triggers release of Ca2+ from endoplasmic reticulum to cytoplasm; and Ca2+ activates various kinases and phosphatases to induce cellular response.1 Intracellular Ca2+ mobilization is monitored by flow cytometry or fluorescence microscopy using fluorescent Ca2+ indicator. The flow cytometer allows us to monitor individual cells that flow through sheath fluid, so we cannot compare the states of each cell before and after stimulation. When cells that are loaded with fluorescent Ca2+ indicator are * To whom correspondence should be addressed. Tel.: +81-761-51-1660. Fax: +81-761-51-1665. E-mail:
[email protected]. † Toyama New Industry Organization. ‡ Japan Advanced Institute of Science and Technology. § These authors contributed equally to this work. | Toyama Medical and Pharmaceutical University. (1) Wienands, J. Immunobiology 2000, 202, 120-133.
8050 Analytical Chemistry, Vol. 77, No. 24, December 15, 2005
analyzed with the flow cytometer, we often observe cells that emit a high intensity of fluorescence without stimulation. These signals become background noise, which consists of 0.1% to sometimes 1% of total cells; thus, it is difficult to monitor Ca2+ mobilization of a minor population of cells whose signals are buried in the noise of a flow cytometer. The fluorescence microscope allows us to observe the states of cells before as well as after stimulation. However, it is difficult to observe signals of a large number of cells under a microscope. Accordingly, it is difficult to monitor Ca2+ mobilization of a minor population of cells. Therefore, it is necessary to construct a microarray platform that can confine a large number of single cells and detect antigenspecific single B-cells before and after stimulation with an antigen from a bulk cell suspension. For high-throughput single-cell separation and analysis, Thorsen et al. reported high-density microfluidic chips that contain plumbing networks with thousands of micromechanical valves and hundreds of individually addressable chambers and showed the separation of single Escherichia coli cells in each chamber.2 To achieve single-cell separation, they diluted cells to create a median distribution of 0.2 cell/compartment, so that reliable capturing of cells in each chamber is difficult. In another recent report, Anderson et al. tested the biomaterial microarrays for their effects on human embryonic stem cell growth and differentiation using populations of human embryonic stem cells.3 However, a single-cell-based assay seemed to be impossible using this microarray format. Also, we recently reported microchamber array and microfluidic chip for measuring high-throughput analysis of cellular fluorescence.4-6 Here, we report an improved microchamber array to monitor Ca2+ mobilization of over 25 000 cells simultaneously at a single-cell level. And we have developed a novel high-throughput screening and analysis system for antigenspecific single B-cells using a microarray, which was carried out by (2) Thorsen, T.; Maerkl, S. J.; Quake, S. R. Science 2002, 298, 580-584. (3) Anderson, D. G.; Levenberg, S.; Langer, R. Nat. Biotechnol. 2004, 22, 863866. (4) Akagi, Y.; Ramachandra Rao, S.; Morita, Y.; Tamiya, E. Science Technol. Adv. Mater. 2004, 5, 343-349. (5) Yamamura, S.; Ramachandra Rao, S.; Omori, M.; Tokimitsu, Y.; Kondo, S.; Kishi, H.; Muraguchi, A.; Takamura, Y.; Tamiya, E. In Proceedings of Micro Total Analysis System (µTAS) 2004; The Royal Society of Chemistry: London, 2004; Vol. 1. pp 78-80. (6) Ramachandra Rao, S.; Yamamura, S.; Takamura, Y.; Tamiya, E. In Proceedings of Micro Total Analysis System (µTAS) 2004; The Royal Society of Chemistry: London, 2004; Vol. 1, pp 61-63. 10.1021/ac0515632 CCC: $30.25
© 2005 American Chemical Society Published on Web 11/12/2005
Figure 1. Schematic novel process for analysis of single B-cell response against antigenic stimulus on a single-cell microarray chip and production of antibody from retrieved single B-cells. B-cells from mouse spleen or human blood were spread on the microarray chip, and singlecell status was achieved under gravitational force. The antigen-specific single B-cells were detected upon stimulation with antigen by monitoring the activated B-cells using the fluorescent calcium indicator, Fluo-4. After retrieval of activated single cells by a micromanipulator under the microscope, RT-PCR amplification of antibody cDNA from each single cell was performed. The ultimate goal of this research is to develop antigen-specific monoclonal antibodies as antibody medicines.
Figure 2. Construction of single-cell microarray chip. (A) LIGA process for the fabrication of single-cell microarray chip. (B, C) SEM images and (D) a real picture of the microarray chip device. The microarray chip is made from polystyrene with over 200 000 microchambers (10-µm width, 12-µm depth, 30-µm pitch). Each microarray chip consisted of 225 (15 × 15) clusters, and each cluster consisted of 900 (30 × 30) microchambers. Each microchamber is cylindrical in shape and can accommodate only a single cell.
detecting antigen-specific single B-cells against an antigen of interest and their retrieval by a micromanipulator for antibody DNA analysis. The single-cell microarray system developed in this study does not need to use myeloma as in the case of a conventional hybridoma technique and can screen the antigen-specific single B-cells directly from the cell suspension and analyze antigen-specific antibody DNA at a single-cell level (Figure 1). This system is simple and easy in its operation and quick enough for making monoclonal antibodies when compared to conventional techniques.
Moreover, this system can perform high-throughput single-cell analysis using chip devices. EXPERIMENTAL SECTION Chip Fabrication. The microarray chip is made from polystyrene with over 200 000 microchambers (10-µm width, 12-µm depth, 30-µm pitch) by using the Lithographie Galvanoformung Abformung (LIGA) process and was performed by Starlight Co. Ltd. (Figure 2). Using X-ray lithography from synchrotron radiaAnalytical Chemistry, Vol. 77, No. 24, December 15, 2005
8051
tion, a poly(methyl methacrylate) (PMMA) as a resist was exposed and patterned with an Au metal mask. After development of the PMMA substrate, the resulting PMMA mold was used for nickel mold construction by electroforming. Finally, polystyrene microarray chip was fabricated from the nickel mold by injection molding (Figure 2A). The microarray surface was rendered hydrophilic by a reactive ion etching (RIE) treatment using the SAMCO RIE system, making it convenient for cell studies. RIE exposure time controlled cell adhesion on the chip surface. Each microarray chip consisted of 225 (15 × 15) clusters, and each cluster consisted of 900 (30 × 30) microchambers. Each microchamber is cylindrical in shape and can accommodate only a single cell. Chip Surface Characterization. The effect of RIE exposure on the microarray chip surface was examined by measuring the contact angle of water on the chip surface using a contact-angle meter (Kyowa Interface Science Co., Ltd.). Cell Preparation. Lymphocytes containing B-cells (60-70%) were extracted from the mashed spleen of the mouse (C3H/He) by using 0.8% (w/v) ammonium chloride solution. Human lymphocytes from peripheral blood were prepared by using FicollConray solution (Immuno-Biological Laboratories Co., Ltd), and B-cells were purified to over 90% using a magnetic cell sorting system (Miltenyi Biotec). Both types of lymphocytes were suspended at 1 × 106 cells/mL in PBS solution (pH 7.4). Single-Cell Detection and Analysis. Lymphocytes from mouse spleen (1 × 106 cells/mL) were mixed with 1 µM fluorescent calcium indicator, Fluo-4-AM (Ex, 494 nm; Em, 516 nm), 1 µM thiol-reactive CellTracker Probe, CellTracker Orange (Ex, 541 nm; Em, 565 nm) (Molecular Probes), or both for 30 min. The lymphocyte suspension was then dispensed manually on to the microarray using a pipet and allowing it to stand for 10 min; cells settle down into the microchambers and on the chip surface under gravitational force. Then the cells that were adhered onto the chip surface were removed by gentle washing with PBS buffer from the edge of the microchamber area using a pipet. The microarray chip was loaded on to the CRBIO IIe-FITC, laser-based fluorescence microarray scanner (Hitachi Software Engineering Co., Ltd.) and scanned. The scanned image was recorded as the image before stimulation. The CRBIO IIe-FITC has a resolution of up to 2.5 µm, sensitivity 10 times)
anti-mouse IgM
anti-human IgM (control)
26650 (100%)
25981 (100%)
18130 (68.0%) 3883 (14.6%) 78 (0.29%)
322 (1.24%) 3 (0.01%) 0 (0%)
a The analysis data showed percentages of single B-cells responding to antigenic stimulus in total lymphocytes on the 32 400 microchamber’ area of the microarry chip.
binding to (Ca2+)i. It is known that the concentration of (Ca2+)i increases after B-cells respond to antigenic stimulation.14 In this experiment, it was observed that major single B-cells on the microarray showed increases in (Ca2+)i after stimulation (Figure 4A). The fluorescence intensity reached to its maximum level after 1-min stimulation, maintained at high magnitude for 2 min, and then decreased gradually (data not shown). Thus, we used 2-min stable duration for analysis. Scattered plot analysis of indvidual cell’s fluorescence intensity (Figure 4B and C, Table 1) enabled us to discriminate B-cells that were activated with anti-mouse IgM antibody. The data from over 30 000 microchambers area in the microarray chip revealed 68% (18 130 cells) of total splenic lymphocytes (26 650 cells) showed more than 2 times higher fluorescence after stimulation with anti-mouse IgM (Figure 4C, Table 1). While cells with over 5 times increase in fluorescence (3883 cells) existed for 14% of total splenic lymphocytes (Figure 4C(I), Table 1). Most of the lymphocytes incubated with control antibody (anti-human IgM) showed less than 5 times increase in fluorescence intensities (Figure 4C(II), Table 1). Analysis of 900 microchambers in the microarray chip also showed the same profile as in the 30 000 microchambers’ data (Figure 4B). Each B-cell expresses membrane-bound antibodies with unique antigen-specificity as antigen receptors. When B-cells were stimulated with an antigen instead of anti-IgM antibody, only a minor population of total B-cells is stimulated and their (Ca2+)i increases. Previous studies reported diverse frequencies in the number of antigen-specific B-cells ranging from 1 in 102 for rabies virus15 to 7 in 105 for myelin basic protein.16 However, it is quite difficult to screen and identify the minor cell population of antigen-specific single B-cells (
